Benefit of Child-to-Parent Kidney Donation


*Corresponding author: Eric P. Cohen,


Use of child-to-parent (CTP) kidney donation may be limited because of ethical concerns as well as doubts about its effectiveness. We used the United Network for Organ Sharing database to examine the effectiveness of CTP kidney donation compared with other types of living-related (LD) kidney donation and to cadaveric kidney donation. Data from 56 873 kidney transplants performed between 1988 and 1998 showed significantly greater transplant and patient survival for CTP kidney transplants compared with cadaveric kidney transplants. The average gain in kidney transplant half-life is 3.6 years for a CTP compared with a cadaveric kidney transplant, and it is estimated that this gain for the recipient far outweighs the 1 in 3000 risk of death to the donor associated with kidney donation. We conclude that CTP kidney donation should not be discouraged, and represents a useful source of transplantable kidneys.


Living donor kidney transplantation provides superior short- and long-term patient and graft survival when compared with cadaveric kidney transplantation (1). Use of living-related donors (LD) has typically been limited to siblings and parents of the recipient. Use of a kidney donated by an adult child to his/her parent has not been widespread, in part because of concerns over the long-term well being of the donor. There may also be an ethical reluctance to place offspring, even adult offspring, at risk for the potential benefit of their parents. A parent or sibling has less expected life years remaining than does a child, even an adult child. The long-term risk of donation is thus much more a concern for child to parent donation than it is for other forms of live donor kidney transplantation.

Limited organ availability has increased pressure on transplant candidates and centers to consider living donation. Within the LD category, kidney transplants from child to parent are a potential source. Two publications from the 1970s show better graft survival in child-to-parent (CTP) kidney transplantation compared with cadaveric kidney transplantation (2,3). Medical care and immunosuppression have changed since then. Thus, it is important to revisit the question of CTP kidney donation in order to evaluate its outcome in the current era.

Materials and Methods

Since October 1, 1987, the United Network for Organ Sharing (UNOS) has administered the Organ Procurement and Transplantation Network (OPTN) in the United States. For this study, OPTN data on 56 873 adult (40 years of age and greater) patients receiving their first kidney transplant, between January 1, 1988, and December 31, 1998, were analyzed. Patients were followed for a minimum of 1 year after transplant. Multiorgan transplants such as kidney pancreas transplants were excluded. This cohort of patients was divided into three groups: Group I, patients who received a CTP kidney; Group II, all other recipients of living donor kidneys; and Group III, recipients of cadaver kidney transplants. Each of the groups was also analyzed by a recipient age group (40–50 years of age, 51–60 years of age, and >60 years of age, respectively).

Demographic categories were compared using the chi-square test. Actuarial graft and patient survival was calculated for each of the three groups using the Kaplan-Meier method. The log-rank test was used to detect differences in the survival curves. Graft and patient half-lives were calculated using Maximum Likelihood Estimates conditioned on 1-year survival and assuming an exponential hazard.

Two types of multivariate analyses were also performed. First, to assess short-term survival, logistic regression analyses were performed to model 1-year graft and 1-year patient survival adjusting for donor, transplant, and recipient variables. Donor and recipient variables included age, gender, and race. Potential transplant variables included type of transplant, cold ischemic time, percent panel reactive antibodies (PRA), HLA matching, and transplant center volume (defined as small: less than 50 transplants per year; medium: 50–100 transplants per year; large: greater than 100 transplants per year). This analysis was repeated censoring for death with a functioning graft. To assess long-term survival, Cox Proportional Hazard models conditioned on 1-year survival were developed. The long-term survival models included the presence or not of acute rejection (defined as the recipient received antirejection medication in the first year post-transplant) as well as the variables included in the short-term analysis. All analyses were performed using SAS, version 8.1 (SAS Institute, Cary, NC).

These studies were approved by the Human Research Review Committee of the Medical College of Wisconsin.


There were 3730 CTP kidney transplants (Group I), 8351 other LD kidney transplants (Group II), and 44 792 cadaver donor kidney transplants (Group III) during the time period analyzed. A summary of recipient demographics for each group is displayed in Table 1. There were significant differences between groups in all demographics shown. For example, and as one would expect, there were more HLA zero mismatches in other living kidney transplants compared with CTP kidney transplants.

Table 1.  Recipient demographics by group: absolute numbers are shown, with percentages in parenthesis.
VariableGroup I
Group II
Other Living
Group III
Recipient Diabetes
 No2702 (72)6230 (75)34582 (77)435140.001
 Yes1028 (28)2121 (25)10210 (23)13359 
Recipient Age Group
 40 − 50963 (26)5237 (63)19613 (44)258130.001
 51 − 601768 (47)2369 (28)16285 (36)20422 
 61 + 999 (27)745 (9)8894 (20)10638 
Recipient Gender
 Female1805 (48)3161 (38)16962 (38)219280.001
 Male1925 (52)5190 (62)27830 (62)34945 
Recipient Race/Ethnicity
 White2384 (64)6252 (75)27228 (61)35864 
 Black781 (21)873 (10)11052 (25)12706 
 Hispanic418 (11)811 (10)4095 (9)53240.001
 Asian80 (2)216 (3)1582 (4)1878 
 Other66 (2)182 (2)797 (2)1045 
 Unknown/Not Reported1 (0)17 (0)38 (0)56 
HLA Mismatch Level
 0111 (3)1741 (21)4358 (10)6210 
 1502 (13)530 (6)1769 (4)2801 
 21352 (36)1287 (15)5656 (13)82950.001
 31612 (43)2013 24)10743 (24)14368 
 401028 (12)11733 (26)12761 
 50984 (12)7239 (16)8223 
 60431 (5)2792 (6)3223 
 Unknown/Not Reported153 (4)337 (4)502 (1)992 
Recipient Peak PRA
 0 − 203351 (90)7521 (90)36595 (82)47467 
 21 − 50211 (6)478 (6)3912 (9)46010.001
 51 + 168 (4)352 (4)4285 (9)4805 
Recipient Peak PRA
Cold Ischemia Time
 Median1122 0.001

The Kaplan-Meier graft survival curves for each group are shown in Figure 1. While the graft survival for patients in the CTP group was significantly lower than that for patients in Group II (other living donor kidneys) (p < 0.001), graft survival for the CTP group was significantly higher than that of Group III (cadaver donor kidneys) (p < 0.001). The same holds true for patient survival, as shown in Figure 2. When the graft survival calculation was censored for death with a functioning graft, there was no difference between Groups I and II, but both groups had superior survival to the survival of Group III (p < 0.001) (Figure 3).

Figure 1.

Kidney transplant actuarial survival, according to donor type. Each curve is significantly different from the others (p < 0.001).

Figure 2.

Patient actuarial survival after kidney trans-plantation, according to recipient group. Each curve is significantly different from the others (p < 0.001).

Figure 3.

Kidney transplant actuarial survival, according to recipient group, and censored for death with a functioning transplant. The curves for CTP kidney and other LD kidney transplants are now superimposed, and not significantly different (p = 0.6), but both are significantly better than the curve for cadaveric transplantation. (p < 0.0001).

Graft half lives by recipient age group were calculated, and these are shown in Figure 4. For each age group, CTP kidneys had a significantly better graft half life than did cadaveric kidneys (all p < 0.001). Although there is a trend in each age group for better graft half lives for other LD kidneys compared with CTP kidneys, this does not attain statistical significance.

Figure 4.

Kidney transplant half-lives, according to recipient age group, and subdivided by donor source. The half lives of cadaveric transplants are significantly lower than those of child-to-parent (CTP) or other living donor transplants in all age groups (p < 0.001 for each comparison). Although the half lives for CTP kidney transplants are numerically less than those of other living donors, this is not statistically significant in any age group.

Child-to-parent and other living donor transplants had better outcome compared with transplants of kidneys from cadaver donors in the short-term multivariate analyses. Cadaveric kidneys had a 28% and 51% higher odds of graft failure at 1 year than did CTP kidneys or other living donor kidneys, respectively, and when censoring for death with a functioning graft the odds of graft failure were significantly higher for cadaver donor kidneys. Cadaver donor kidneys were associated with a 21% and 51% higher odds of recipient death at 1 year than CTP and other living donor kidneys, respectively. All of the significant risk factors for the multivariate logistic regression analyses are listed in Tables 2–4. The donor and recipient ages are shown in Table 5. The younger age of the CTP donors is evident. Six CTP donors were under the age of 18 years.

Table 2.  Logistic regression – 1 year graft survival significant variables (p < 0.05)
DonorOther Living Donor vs Cadaver Transplant0.491(0.443,0.543)
 Child to Parent Living Donor vs Cadaver Donor Transplant0.724(0.635,0.825)
 Donor Age  
  45 vs 351.169(1.161,1.177)
  55 vs 351.529(1.505,1.550)
 Female vs Male1.122(1.066,1.182)
 Black vs White, Asian & Other1.311(1.212,1.418)
 Hispanic vs White, Asian & Other1.219(1.109,1.340)
RecipientAsian vs White & Black0.695(0.594,0.813)
 Other Race vs White & Black0.719(0.586,0.882)
 Hispanic vs White & Black0.730(0.663,0.805)
 Peak PRA1.006(1.005,1.007)
 Years of Dialysis Prior to Transplant1.042(1.034,1.050)
 51−60 Years vs 40−50 Years1.158(1.094,1.226)
 61+  Years vs 40−50 Years1.387(1.296,1.485)
TxLarge Volume Center vs Medium and Small Volume0.866(0.818,0.916)
 Centered Cold Time (Cad Mean = 22.6 hrs Living Mean = 1.45 hrs)1.008(1.005,1.011)
 1 HLA Mismatch vs 0 Mismatches1.271(1.090,1.482)
 2 HLA Mismatches vs 0 Mismatches1.422(1.266,1.598)
 3 HLA Mismatch vs 0 Mismatches1.364(1.225,1.518)
 4 HLA Mismatches vs 0 Mismatches1.601(1.439,1.781)
 5 HLA Mismatches vs 0 Mismatches1.738(1.554,1.945)
 6 HLA Mismatches vs 0 Mismatches1.922(1.679,2.199)
 HLA Mismatch Level Unknown vs 0 Mismatches1.646(1.314,2.062)
 Transplant Year0.924(0.916,0.932)
Table 3.  Logistic regression – 1 Year patient survival significant variables (p < 0.05)
DonorChild to Parent Living Donor vs Cadaver Donor Transplant0.794(0.674,0.934)
 Other Living Donor vs Cadaver Transplant0.488(0.432,0.564)
 Donor Age  
  45 vs 351.123(1.112,1.135)
  55 vs 351.349(1.321,1.377)
 Hispanic vs White, Asian & Other1.246(1.096,1.417)
 Black vs White, Asian & Other1.297(1.159,1.452)
RecipientHispanic vs White & Other0.696(0.607,0.799)
 Asian vs White & Other0.767(0.618,0.952)
 Black vs White & Other0.854(0.782,0.933)
 Female vs Male0.884(0.821,0.951)
 Years of Dialysis Prior to Transplant1.058(1.048,1.069)
 51−60 Years vs 40−50 Years1.528(1.408,1.658)
 61+  Years vs 40−50 Years2.316(2.116,2.534)
 Peak PRA1.004(1.002,1.005)
TxLarge Volume Center vs Small & Medium Volume0.837(0.774,0.905)
 Centered Cold Time (Cad Mean = 22.6 hrs Living Mean = 1.45 hrs)1.006(1.002,1.009)
 6 HLA Mismatch vs 0,1,2,3,4 Mismatches1.218(1.057,1.404)
 5 HLA Mismatches vs 0,1,2,3,4 Mismatches1.204(1.097,1.321)
 Transplant Year0.939(0.928,0.950)
Table 4.  Logistic regression – 1 year graft survival – censoring for death with functioning kidney – significant variables (p < 0.05)
DonorOther Living Donor vs Cadaver Transplant0.495(0.439,0.559)
 Child to Parent Living Donor vs Cadaver Donor Transplant0.620(0.522,0.738)
 Donor Age  
  45 vs 351.204(1.194,1.213)
  55 vs 351.668(1.644,1.692)
 Female vs Male1.143(1.075,1.216)
 Hispanic vs White, Asian & Other1.173(1.047,1.314)
 Black vs White, Asian & Other1.363(1.243,1.493)
RecipientOther Race vs White & Black0.662(0.513,0.854)
 Asian vs White & Black0.699(0.582,0.841)
 Hispanic vs White & Black0.727(0.648,0.815)
 Peak PRA1.008(1.007,1.009)
 Years of Dialysis Prior to Transplant1.033(1.024,1.043)
TxLarge Volume Center vs Medium & Small Volume0.855(0.799,0.916)
 Centered Cold Time (Cad Mean = 22.6 hrs Living Mean = 1.45 hrs)1.009(1.006,1.012)
 1 HLA Mismatch vs 0 Mismatches1.446(1.194,1.751)
 2 HLA Mismatches vs 0 Mismatches1.649(1.424,1.910)
 3 HLA Mismatches vs 0 Mismatches1.591(1.388,1.823)
 4 HLA Mismatches vs 0 Mismatches1.992(1.741,2.279)
 5 HLA Mismatches vs 0 Mismatches2.124(1.845,2.446)
 6 HLA Mismatches vs 0 Mismatches2.380(2.018,2.808)
 HLA Mismatch Level Unknown vs 0 Mismatches2.129(1.642,2.759)
 Transplant Year0.909(0.900,0.919)
Table 5.  Conditional Cox's proportional hazards regression analysis long-term graft survival conditioned on 1 year survival
DonorOther Living Donor vs Cadaver Transplant0.640(0.596,0.687)
 Child to Parent Living Donor vs Cadaver Donor Transplant0.818(0.746,0.897)
 Donor Age  
  45 vs 351.158(1.151,1.164)
  55 vs 351.434(1.417,1.452)
 Black vs White, Hispanic, Asian & Other1.190(1.118,1.266)
 Female vs Male1.041(1.000,1.082)
RecipientFemale vs Male0.849(0.816,0.884)
 Black vs White, Hispanic, Asian & Other1.596(1.526,1.669)
 Peak PRA1.001(1.000,1.002)
 Years of Dialysis Prior to Transplant1.025(1.018,1.031)
 51−60 Years vs 40−50 Years1.066(1.021,1.113)
 61+  Years vs 40−50 Years1.406(1.335,1.481)
TxLarge Volume Center vs Small Volume0.853(0.812,0.896)
 Medium Volume Center vs Small Volume0.950(0.909,0.994)
 Recipient Received Anti-rejection Medication in the First Year Post-Tx1.422(1.340,1.509)
 6 HLA Mismatch vs 0,1,2,3,4,5 Mismatches1.190(1.097,1.290)
 Transplant Year0.977(0.969,0.986)

The risk of long-term graft failure was 18% higher for cadaver donor kidneys than CTP kidneys. There was also a 12% greater risk of death beyond 1 year for cadaver kidney recipients compared with recipients of CTP kidneys. Cadaver donor kidneys were also associated with a 36% higher risk of long-term graft failure and 32% higher risk of death than other living donor kidneys. All of the significant risk factors for the Cox Proportional Hazard models are listed in Tables 6 and 7.

Table 6.  Conditional Cox's proportional hazards regression analysis long-term patient survival conditioned on 1 year survival
DonorOther Living Donor vs Cadaver Transplant0.678(0.619,0.741)
 Child to Parent Living Donor vs Cadaver Donor Transplant0.882(0.792,0.984)
 Donor Age  
  45 vs 351.111(1.104,1.118)
  55 vs 351.301(1.283,1.318)
 Black vs White, Hispanic, Asian & Other1.110(1.023,1.203)
RecipientOther vs White0.772(0.638,0.935)
 Hispanic vs White0.777(0.707,0.855)
 Asian vs White0.805(0.684,0.947)
 Black vs White1.152(1.085,1.223)
 Female vs Male0.814(0.774,0.857)
 Years of Dialysis Prior to Transplant1.043(1.035,1.051)
 Peak PRA1.001(1.000,1.003)
 51−60 Years vs 40−50 Years1.427(1.350,1.507)
 61+  Years vs 40−50 Years2.297(2.158,2.444)
TxLarge Volume Center vs Small Volume0.852(0.801,0.906)
 Medium Volume Center vs Small Volume0.929(0.878,0.983)
 Transplant Year0.975(0.965,0.986)
 Recipient Received Anti-rejection Medication in the First Year Post-Tx1.140(1.051,1.237)
 6 HLA Mismatches vs 0,1,2,3,4,5 Mismatches1.126(1.014,1.251)
Table 7.  Death with functioning graft (DWFG), cause of death, recipient and donor ages
Cause of death      
Unknown/Not Reported15932.422032.1315145.0
Recipient Age   
Mean Age55.749.553
Median Age564852
Age Range417941794188
Donor Age   
Mean Age30.34434.2
Median Age304433
Age Range14701281<189


The dominant form of kidney transplantation remains cadaveric transplantation. Improvements in use of immunosuppression and medical care have been associated with improved patient and graft survival in recent years (4). Specifically, the half-life for a cadaveric graft has increased from 11 years in 1988 to 19.5 years in 1995. The impact of these improvements is limited by the supply of donated cadaver organs. Thus, alternatives to cadaveric transplantation remain important. The use of living donor kidneys, the major alternative to cadaveric transplantation, has more than doubled during the past decade (5). Specifically, of 8657 kidney transplants carried out in 1989, 1903 were from living donors, whereas of 13 290 kidney transplants performed in 1999, 5227 were from living donors. The present paper shows that CTP kidney donation should not be ignored as a useful source of living-donor kidneys for transplantation.

The superior graft and patient survival of living donor compared with cadaveric kidney transplants has led to the use of multiple sources of living donor kidneys. Not only sibling-to-sibling donation, but also parent-to-child, other relative, and emotionally related donation are now commonly used. Schemes for unrelated-donor kidney sharing have been suggested as ways to further enhance the number of kidney transplants. Commercial donation is also used in some parts of the world, despite ethical and medical worries (6).

Use of CTP donation may not be quite as controversial as is commercial kidney donation, but it nonetheless raises some concerns. While donation of renewable tissues, such as blood or bone marrow, is only a transient inconvenience for the donor, the issues of short- and long-term safety are very real for the living kidney donor. A recent large survey published in 1995 showed 1 in 3000 mortality from kidney donation (7). A similar mortality risk is cited in a 1992 study (8). Neither of these reports identifies the CTP donor specifically.

In addition to the short-term risks of kidney donation, there are potential longer-term risks, although several studies indicate that these risks are minimal. Anderson and colleagues showed that 100 kidney donors had stable and virtually normal kidney function for 10–20 years of follow up, with no increase in the frequency of hypertension compared with the general population (9). Williams and colleagues compared the long-term renal function of 38 kidney donors with their siblings at an average of 13 years after donation (10). They found some elevation in urinary protein excretion, on average to 200 mg/day, in the donors compared with their nonrecipient siblings. The serum creatinine levels were 1.2 mg/dL in the donors as compared with 1.0 mg/dL in their nonrecipient siblings, but the donors had no excess of hypertension. Robitaille and colleagues studied 27 subjects who had undergone uninephrectomy in childhood and found no significant effect on kidney function, proteinuria, or blood pressure at an average of 23 years of follow up (11). A recent report suggests that living donors, with more than 20 years of follow up, may be at some risk for kidney disease, but this has not been confirmed (12). Thus, the long-term risks of kidney donation appear to be acceptable.

There is an acknowledged morbidity for the donor, in terms of pain, possible wound infection, and other operative complications (13). This risk to the donor should be weighed against the morbidity of long-term dialysis for the recipient. By way of an example, an average chronic dialysis patient in the United States can expect 1.5 hospitalizations per year, whereas the average transplant patient can expect only 0.5 hospitalization per year (1). As donor and recipient will both have been hospitalized for the donation/transplant, after 2 years the total hospitalization risk (a crude measure of morbidity) will be in favor of the CTP option when compared with dialysis. The use of laparoscopic surgery for kidney donation may further tilt this estimate in favor of CTP transplantation compared with chronic dialysis.

Parental concern for offspring could deter CTP kidney donation. This is enhanced by the reluctance of the physician or surgeon to recommend or perform a procedure for kidney donation that might cause illness or death. That reluctance may be enhanced if the physician or surgeon is a parent. However, and as discussed by Hou, the nonmedical public believes in autonomy, which means that the potential donor believes that (s)he should have the final decision (6). That decision must, nonetheless, be made based on existing data.

The data presented here show a distinct and significant superiority for graft and patient survival rates for CTP kidney transplants as compared with cadaveric kidney transplants. The logistic regression and Cox analyses confirm this better survival, even after adjustment for important factors such as HLA matching, cold ischemia time, time on dialysis, years of transplantation, and diabetic status. The apparent superiority of other living donor kidneys, when compared with CTP kidneys, vanishes when the data are censored for death with a functioning graft, which suggests a higher nonimmunological mortality in the CTP recipients compared with the recipients of other LD kidneys. This is shown in Table 5, which shows the greater percentage of cardiovascular deaths in the CTP recipients. It is possible that unidentified confounding variables may account for these differences. Still, the greater average age of the CTP recipients compared with other LD recipients is likely to account for their greater occurrence of death with a functioning graft. Nonetheless, a better graft half-life in CTP kidneys compared with cadaveric kidneys is evident in each recipient age group (Figure 4).

These data suggest that CTP kidneys may not have quite as good an outcome as do other LD kidneys. Often, however, the choice faced by the patient and transplant team is between a CTP kidney and a cadaveric kidney. It is useful then to estimate the potential gain afforded by CTP as opposed to cadaveric kidney transplantation. A 1 in 3000 risk of death related to kidney donation translates to 1.3 deaths in the ∼ 4000 CTP donors of this study. With an average CTP donor age of 30 years in this study, one would expect 45 more years of life for an average donor, and thus 60 years of life lost as a result of the 1.3 deaths. The patient half-life per case is estimated to be 13.5 years for a cadaveric kidney recipient, as opposed to 17.1 years for a CTP recipient, providing for 13 400 years of life gained by CTP over cadaveric transplantation in the cohort of 3730. Therefore, with CTP kidney donation, the theoretical gain in life is far greater than the life lost. In the case of a kidney transplant to diabetic recipients aged 60 years and more, this gain in life years is narrower, but still apparent. Theoretically there are approximately 4 years of life lost in the 267 donors of this group. While the gain in patient half life for a CTP as compared with a cadaveric kidney in this group is not statistically significant, it is 1.8 years, which in the entire cohort of 267 is 480 theoretical years of life gained with a CTP compared with a cadaveric kidney in the older diabetic recipient.

On balance, then, CTP kidney donation cannot be discouraged on objective grounds. While some family members or transplant team members may have ethical concerns, the overall data are strongly in favor of a superior outcome of CTP compared with cadaveric transplantation. It is possible that an older, sicker potential recipient may not derive significant benefit from CTP transplantation, or even from kidney transplantation at all, so the decision for kidney transplantation must remain individualized.


These data were presented at the Transplantation 2001 meeting in Chicago, IL, and have been published in abstract form. Eric Cohen initiated this study and wrote the paper. John Rosendale and Christine Bong performed the statistical analyses and drafted the tables and figures. Sundaram Hariharan planned the data collection, provided data interpretation, and revised the paper.